Der Open-Access-Publikationsserver der ZBW – Leibniz-Informationszentrum Wirtschaft
The Open Access Publication Server of the ZBW – Leibniz Information Centre for Economics
Hess, Dieter; Immenkötter, Philipp
Working Paper
Optimal leverage, its benefits, and the
business cycle
CFR working paper, No. 11-12
Provided in cooperation with:
Universität zu Köln
Suggested citation: Hess, Dieter; Immenkötter, Philipp (2011) : Optimal leverage, its benefits,
and the business cycle, CFR working paper, No. 11-12,
Die ZBW räumt Ihnen als Nutzerin/Nutzer das unentgeltliche,
räumlich unbeschränkte und zeitlich auf die Dauer des Schutzrechts
beschränkte einfache Recht ein, das ausgewählte Werk im Rahmen
der unter
nachzulesenden vollständigen Nutzungsbedingungen zu
vervielfältigen, mit denen die Nutzerin/der Nutzer sich durch die
erste Nutzung einverstanden erklärt.
Leibniz-Informationszentrum Wirtschaft
Leibniz Information Centre for Economics
Terms of use:
The ZBW grants you, the user, the non-exclusive right to use
the selected work free of charge, territorially unrestricted and
within the time limit of the term of the property rights according
to the terms specified at
By the first use of the selected work the user agrees and
declares to comply with these terms of use.
CFR Working Paper NO. 1111-12
Optimal Leverage, its Benefits, and
the Business Cycle
D. Hess • P. Immenkötter
Optimal Leverage, its Benefits,
and the Business Cycle∗
Dieter Hess†, Philipp Immenk¨otter‡
We study the effect of the business cycle on optimal capital structure choice and the
benefit to leverage. We propose a regime switching model with a state-dependent
cash flow process to capture macroeconomic risk in a firm’s cash flow. Our model
is parsimonious but still realistic and allows for a wide range of analysis. We find
pro-cyclical optimal leverage ratios, benefits to leverage, and costs of operating at a
non-optimal leverage. If macroeconomic risk decreases, i.e. earnings become more
stable and growth rates less volatile, optimal leverage and its benefits increase due
to lower default risk. The regime switching property of EBIT traces observed EBIT
paths closely and is applicable to a wide range of corporate valuation models. Our
model offers novel empirically testable implications, such as higher tax benefits after
the change in macroeconomic risk since the late 1980s and common capital structure
adjustments in recessions and around turning points.
Keywords: capital structure; macroeconomic risk; regime switching; benefit to leverage
JEL classification: E44; G12; G32
We are grateful for valuable discussions and comments from Christian Hellwig, Alexander Ludwig, Anna
auser, Martin Ruckes, Marliese Uhrig-Homburg, the participants of the research seminars in Cologne
and Karlsruhe (KIT), the 14th meeting of the Swiss Society for Financial Market Research (SGF), and the
47th annual meeting of the Eastern Finance Association (EFA).
University of Cologne, Corporate Finance Seminar and CFR Cologne
Corresponding author, University of Cologne, Corporate Finance Seminar, contact information:
[email protected], Tel. +49-221-4707875.
The business cycle is essential for understanding corporate financing decisions. To analyze
how the macroeconomic risk affects optimal leverage ratios, we propose a structural model
of optimal capital structure that incorporates changing macroeconomic conditions through
the firm’s cash flow channel. Our model is parsimonious but at the same time realistic
and allows for a wide range of analysis. We show that optimal leverage and its benefits
vary pro-cyclically, and a reduction of macroeconomic risk lowers optimal leverage ratios
but influences benefits to leverage only marginally. The costs that a corporation faces if it
operates at a non-optimal leverage are higher in recessions.
In contrast to previous approaches, we model earnings before interest and taxes (EBIT)
with a stochastic process that depends on the business cycle. In expansions EBIT follow
a positive trend while in contractions they decrease on average. The turning points of the
economy are determined stochastically by a Markov chain. Following Goldstein, Ju, and Leland (2001) and the trade-off theory, the firm chooses its optimal financing mix by balancing
tax benefits and default costs which in turn depend on the macroeconomic conditions. As
a result we find that optimal capital structures and benefits to leverage strongly differ in
expansions and recessions.
Our model shows that optimal leverage choice varies pro-cyclically with the business
cycle. In expansions firms choose a higher amount of debt for financing their investments,
while they turn to equity financing in contractions. Positive growth expectations decrease
a firm’s default risk and increase its debt capacity. In contrast, default is more likely in
recessions and the firm behaves optimal by choosing a higher amount of equity to reduce
default risk.
The benefit to leverage, defined as the ratio of the levered and unlevered firm value, is
pro-cyclical as well. Estimating the parameters of our model to reflect S&P500 firms, we find
that by issuing debt the unlevered asset value of the firm is increased by 5% in expansions
and by 4% in contractions. Despite increasing default risk, tax shields are important means
to maximize shareholders wealth in contractions. In expansions the levered firm value is 23%
higher than in recessions, but the benefits increase only by 1%. Hence, benefits account for
only a small fraction of the gain in the levered firm value.
If managers want to determine their firm’s optimal leverage ratios, they need to precisely
assess the present state of the economy and the expected growth rate of EBIT. If they fail to
identify the present conditions, they come to a non-optimal leverage choice. Firms operating
at a non-optimal leverage ratio face costs of being over- or underlevered. We find that
these costs are higher in recessions than in expansions. In recessions marginal default costs
increase more rapidly with outstanding debt which makes being overlevered more costly in
this state. If managers issue too little debt in a contraction, then they miss a substantial
amount of tax benefits which are an important way to create shareholder value since earnings
will decline on average. For a small deviation from the optimum, the costs are only a small
fraction of the levered firm value, but if the optimum changes due to a switch of the state,
then the firm faces high costs. If capital structure adjustments are costly, then the firm
adjusts their leverage more often at turning points of the economy, because there it is more
likely that the increase in value exceeds the adjustment costs. Besides this finding, leverage
adjustments should generally be more common in recessions because the costs of being overor underlevered are higher in this state.
As observed by Stock and Watson (2002), macroeconomic risk has changed over time.
After the 1980s, recessions became milder and economic growth less volatile. This shift
in macroeconomic conditions affects corporate financing policies. Our model shows that
mild recessions and less volatile growth rates lead to higher levered firm values, because the
loss of cash flows in recessions is smaller. As a response optimal leverage increases due to
a reduction in default risk. However, the benefits to leverage are hardly affected by the
changing conditions. The fraction of the levered firm value that corresponds to the benefits
to leverage varies only slightly. If firms behave optimal they can keep the benefits to leverage
at the same level, independent of the state of the economy.
Our theoretical results on pro-cyclical leverage ratios explain recent empirical findings by
Korteweg (2010) who assesses a significant positive difference in optimal leverage ratios over
the business cycle. Our findings are also in line with Covas and Den Haan (2010) who come to
the result of pro-cyclical debt issuances. In their empirical study, Korajczyk and Levy (2003)
analyze observed leverage ratios on the basis of financial constraints and state that financially
unconstrained firms have countercyclical leverage ratios. Our model addresses optimal rather
than observed leverage ratios, which are not necessarily identical. For example, according
to the trade-off theory a profitable firm operates at a high optimal leverage ratio. But over
time high retained earnings decrease the observed leverage constantly, so that observed and
optimal ratios do not correspond to each other. In our model optimal leverage is not affected
by the historic outcomes, it reflects the optimal leverage choice at a certain point in time.
The main distinction of our model to previous approaches is that the firm faces unlimited
downward risk once a recession sets in. In constrast, modeling EBIT with a geometric
Browian motion multiplied with a state variable that switches between two values to reflect
macroeconomic risk (e.g. Hackbarth, Miao, and Morellec (2006)) implies that the economic
outlook at the beginning of a recession is rather bright. Once the firm has survived the
jump down to a much lower EBIT level at the start of a recession, the upside potential from
jumping back to the old level is huge. This leads to the counterintuitive implication that
the economic outlook at the beginning of a recession is even brighter than in the middle of
an expansion. Hence, the assumption that drift rates rather than the level of EBIT switch
is the central distinction in our model. It is mainly responsible for the fact that we find
pro-cyclical rather than counter-cyclical leverages to be optimal.
Two independent, recent papers construct a similar framework but have a different focus. Chen (2010) uses regime switching processes to model a firm’s cash flow, the outcome
of the economy, and the consumer price index. His analysis focuses on credit spreads and
default rates Bhamra, Kuehn, and Strebulaev (2009) combine a structural model with a
consumption-based asset pricing model to explain leverage ratios at aggregated and indi-
vidual levels. Their approach integrates the effect of the business cycle in the aggregate
consumption and through an additional systematic volatility component. In our analysis
we examine the optimal leverage choice, the benefit to leverage, costs of being over- or underlevered and changing macroeconomic conditions. Our approach focuses on the effect of
state dependent EBIT, not of the consumption in the economy. Moreover, our approach is
more parsimonious and we are able to trace the observed effects closely back to our state
dependent earnings process.
The rest of the paper is organized as follows: In section 2 we define the economy of the
model and in section 3 we use the contingent claims valuation technique to derive analytical
functions for the levered firm and debt value. In section 4 we estimate the model’s parameters and analyze implications on optimal leverage, the benefit to leverage, and changing
macroeconomic conditions. Finally, section 5 concludes. Derivations are contained in the
The model
We consider a continuous time economy with a representative firm and two possible states
of the economy, i.e., expansion (i = 1) or recession (i = 2). All agents know the present
state i0 at all times and its characteristics. The transition from one state to the other is
given by a Markov chain Mt with Poisson transition probabilities. The rate of leaving the
present state i in the infinitesimal time interval dt is denoted λi . All agents are risk neutral
and discount cash flows with the constant risk free rate r. Corporate earnings are taxed at
the constant tax rate τ and management acts in best interest of shareholders.
Upon this economy, we consider an infinitely-lived firm whose assets generate stochastic
EBIT xt . Because of different earning perspectives in different states of the economy, xt
follows a geometric Brownian motion that changes its drift and volatility components at
turning points of the economy, called regime-switching-process:1
dxt = µi xt dt + σi xt dWt ,
µ1 , µ2 ∈ R,
where i = 1, 2
σ1 , σ2 > 0,
x0 > 0
Uncertainty is modeled through the complete probability space (Ω, F, P), where the σalgebra F is generated by the Markov chain Mt and standard Brownian motion (Wt )t≥0 .
The probability measure P is the product of the distribution of dWt and Mt . To obtain a
realistic setting of EBIT across the business cycle, we assume µ1 > 0 and µ2 < 0, i.e. the
expected earnings growth is positive in an expansion and negative in a recession. The risk of
deviation from the present EBIT-trend µi is reflected through the state dependent volatility
component σi . The present value of expected perpetually generated after-tax EBIT is the
value of the firm’s unlevered assets u(x).
Since the firm’s EBIT is subject to taxation at the constant tax rate τ , the firm has
incentives to issue debt in order to generate tax benefits and to create a levered firm value
v(x) exceeding the unlevered asset value u(x). We consider a stationary debt environment
according to Leland (1998) where a firm initially issues a certain amount of debt with principal P , fair coupon C. Throughout time a constant fraction m of outstanding debt is retired
continuously and replaced by the same amount of debt with the same coupon and retirement
rate, so that the outstanding amount of debt P is constant over time. Debt is issued at par
so that the principal P equals the initial market value of debt d(x0 ) and the firm has to
pay the fair coupon that incorporates the risk of the volatile EBIT process. The value of
shareholders equity e(x) is the residual claim to the levered firm value v(x) after subtraction
of the market value of debt d(x).
The issuance of debt bears default risk. We incorporate an endogenous default decision
according to Leland and Toft (1996), where the decision to default belongs to the shareholders
and default is triggered if EBIT xt falls beneath state dependent default thresholds, K1 and
For details on regime-switching-processes see Guo (2001)
K2 . The thresholds correspond to the point where shareholders optimally stop injecting funds
into the firm because the cost of remaining active equals its benefits. Since the benefits are
state dependent and higher in an expansion due to the positive difference in the growth rates
µ1 −µ2 , we find that K1 < K2 . In words, the firm will default earlier in a recession than in an
expansion. When default occurs, bondholders receive the unlevered asset value less default
costs and equity becomes worthless. Default costs are reflected through a state dependent
recovery rate 0 < αi ≤ 1, so that the payment to bondholders in case of default corresponds
to d(x) = αi u(x), i denoting the present state at time of default.
All claims, such as the levered firm value v(x), the unlevered asset value u(x), debt value
d(x), and equity value e(x) are state dependent functions of EBIT xt , which is denoted by
the index i:
v(xt ) =
v1 (xt )
if it = 1
v2 (xt )
if it = 2
Initially managers choose the outstanding amount of debt P to maximize the levered firm
value v(x) = vi0 (x) and hence the shareholders wealth. The maximum is achieved when the
marginal tax benefits equal the default cost. Since the value of the levered firm v(x) depends
on the present state, the maximization problem has two solution, namely for each possible
present state one solution:
max vi (x) i = 1, 2
We define leverage as the ratio of debt and the levered firm value li = di (x)/vi (x). l1 is
the leverage in an expansion, l2 in a recession, and li∗ is the optimal ratio that maximizes
the levered firm value in the corresponding state i. The benefit to leverage is defined as
bi =
vi (x)−ui (x)
ui (x)
and refers to the present value of the expected future benefits of issuing debt
in t = 0. Since the firm value can be split up into the value of unlevered assets and the
benefits, the maximization of the levered firm value is equivalent to the maximization of the
Valuation of corporate securities
In this section we construct a static contingent claims valuation model with the basic
characteristics of Hackbarth, Miao, and Morellec (2006) to derive closed form solutions for
the value of the firm’s unlevered asset value, the levered firm value, its outstanding amount
of debt, and the value of shareholder’s equity. All claims are denoted as functions of EBIT
xt and dependent on the state i = 1, 2.
If the default threshold in expansions K1 is lower than in recessions K2 , we have three
disjoint regions that can be analyzed separately. We define the region x ≤ K1 as the default
region. Here, the firm is liquidated in both states. Second, K1 < x ≤ K2 , defines the
transient region where the firm is active in state one, but liquidated in state two. Third, the
action region K2 < x, where the firm is active in both states. As shown in section 4, the
case of K1 < K2 is sufficient and hence we will not discuss other scenarios.
Unlevered asset value
Following Mello and Parsons (1992) the value of unlevered assets u(x) corresponds to the
present value of a perpetual claim to after-tax EBIT.
u(x) = E
(1 − τ )xs ds x0 = x, i0 = i ,
i = 1, 2
It can be compared to the value of the firm that does not issue debt. An application of Itˆo’s
Lemma to each state of the economy yields a set of ordinary differential equations for the
value of unlevered assets ui :2
1 2 2 00
x σ1 u1 + µ1 xu01 − ru1 + λ1 (u2 − u1 ) + (1 − τ )x = 0
1 2 2 00
x σ2 u2 + µ2 xu02 − ru2 + λ1 (u1 − u2 ) + (1 − τ )x = 0
The equation consists of three parts: first, a linear Black-Scholes operator 12 x2 σi2 u00i + µi xu0i −
rui captures the change of the unlevered firm value due to the movement of the Brownian
motion. The second part λi (uj − ui ) is the change in values arising from a regime shift
multiplied by the rate of leaving state i. The third part (1 − τ )x represents the perpetual
claim to after-tax EBIT.
Under the boundary conditions
ui (x)
lim ui (x) < ∞ and lim
the solution to (5) and (6) is
u1 (x) = w1 · x
u2 (x) = w2 · x .
with the constants w1 and w2 given by
(1 − τ )(µ2 − r − λ1 − λ2 )
(µ1 − r − λ1 )(µ2 − r − λ2 ) − λ1 λ2
(1 − τ )(µ1 − r − λ1 − λ2 )
w2 = −
(µ1 − r − λ1 )(µ2 − r − λ2 ) − λ1 λ2
w1 = −
The unlevered asset value is a linear function of the present level of EBIT. The constant
coefficients wi incorporate the growth rate in the present state µi and the possible switch
to the other state with a different growth rate. Intuitively, for the considered scenario with
µ1 > µ2 , we find that u1 (x) > u2 (x). Moreover, the unlevered asset value is independent of
The derivation of the differential equations is contained in the appendix A.1.
the volatility σi of EBIT because it is the expected value of the process, reflecting the future
drift not the volatility.
Value of corporate debt
The state dependent value of all outstanding corporate debt di (i = 1, 2) is the present
value of the continuous coupon payment C and retirement of the principal mP as long as
the firm is solvent. Upon default bondholders receive the state dependent unlevered asset
value diminished by the recovery rate αi . In the default region the firm is liquidated in both
states which leads to a debt value of αi wi x. In the transition region (K1 < x ≤ K2 ) the
firm is active in state one but defaults in state two. Therefore, the value of outstanding debt
in state two is d2 (x) = α2 w2 x. An application of Itˆo’s lemma3 in state one yields a set of
differential equations for d1 (x):
1 2 2 00
x σ1 d1 + µ1 xd01 − (r + m)d1 + λ1 (α2 w2 x − d1 ) + C + mP = 0 .
The structure of the equation matches (5) and (6). Note that in the transient region α2 w2 x
corresponds to d2 (x) and the constant payment C + mP is the absolute outstanding coupon
plus the retirement of the principal.
In the action region (K2 < x) the firm is active in both states. Itˆo’s lemma yields in
analogy to (5) and (6) a set of differential equations:
1 2 2 00
x σ1 d1 + µ1 xd01 − (r + m)d1 + λ1 (d2 − d1 ) + C + mP = 0
1 2 2 00
x σ2 d2 + µ2 xd02 − (r + m)d2 + λ2 (d1 − d2 ) + C + mP = 0
In order to obtain continuous solution functions for di (x), we use continuity conditions at
The exact derivation is contained in the appendix A.2.
x = K1 and x = K2 :
d1 (K1 ) = α1 w1 K1
d2 (K2 ) = α2 w2 K2
lim d1 (x) = lim+ d1 (x)
In addition, at x = K2 we require a smooth fit of d1 (x) to obtain a C 1 -function outside of
the default region.
lim− d01 (x) = lim+ d01 (x) .
Solving (11), (12), (13) subject to (14) - (17), we receive explicit functions for the value of
state dependent corporate debt:
Theorem 1: Value of corporate debt
If a firm’s EBIT is given by (1), its debt structure by (C, m, P ), and the default policy
by K1 < K2 , then the value of corporate debt d(x) is state dependent and satisfies:
d1 (x) =
α1 w1 x
for x ≤ K1
A1 xγ1 + A2 xγ2 + r+m+λ
A3 xβ1 + A4 xβ2 + C+mP
α2 λ1 w2 x
r+m+λ1 −µ1
for K1 < x ≤ K2
for K2 < x
d2 (x) =
α2 w2 x
for x ≤ K2
b 3 A 3 x β 1 + b 4 A 4 x β 2 +
for K2 < x
where β1 , β2 < 0, γ1 > 0, γ2 < 0, b3 < 0, b4 > 0 and A1 , A2 , A3 , A4 are the coefficients determined by the boundary conditions (14) - (17). The derivation and explicit
formulas of the exponents and coefficients are contained in the appendix A.2.
In the action region the value of corporate debt consists of two parts: First,
is the risk-
free value of perpetual debt. Second, the negative sum A3 xβ1 + A4 xβ2 reflects the discount of
the risk-free value due to default risk and a possible regime shift. In state one the default risk
can increase because the drift can switch from µ1 > 0 to µ2 < 0. In contrast, a switch from
state two to state one decreases default risk because on average EBIT x moves further away
from the default threshold Ki due to the positive drift µ1 . With increasing x the exponential
terms vanish and the whole expression converges to the value of risk-free debt
the structure of the differential equations, we can express the discount due to default risk
in an recession by multipyling the singel terms of the default risk in an expansion by b3 and
b4 respectivly. In the transient region (K1 < x ≤ K2 ) a switch from state one to state two
results in a sudden default which is reflected by the increased discount rate r + m + λ1 − µ1
of the perpetual debt claim. The additional term
α2 λ1 w2 x
r+m+λ1 −µ1
incorporates the default value
in state two. The analytical formulas extend those of Leland (1994) by the terms of the
transient region and the different exponents in the transient and action region.
Levered firm value
The value of the levered firm is the present value of a claim to after tax EBIT plus the
tax shield as long as the firm is solvent. In default it is the liquidation value less default
costs. It can be treated as the solution to a system of differential equations constructed in
the same way as in the case of corporate debt. Because of the different growth perspectives
of EBIT in different states, the levered firm value vi (x) (i = 1, 2) is state dependent as well.
In the transient region the firm defaults in state two but remains active in state one.
Hence, we have a single equation for the value of the levered firm in the transient region
(K1 < x ≤ K2 ):
1 2 2 00
x σ1 v1 + µ1 xv10 − rv1 + λ1 (α2 w2 x − v1 ) + (1 − τ )x + τ C = 0 .
In the action region the firm is active in both states and satisfies the equations
1 2 2 00
x σ1 v1 + µ1 xv10 − rv1 + λ1 (v2 − v1 ) + (1 − τ )x + τ C = 0 ,
1 2 2 00
x σ2 v2 + µ2 xv20 − rv2 + λ1 (v1 − v2 ) + (1 − τ )x + τ C = 0 .
The respective continuity and smoothness conditions read
v1 (K1 ) = α1 w1 K1 .
v2 (K2 ) = α2 w2 K2 .
lim v1 (x) = lim+ v1 (x) ,
lim− v10 (x) = lim+ v10 (x) .
Solving (20) - (22) subject to (23) - (26) gives theorem 2:
Theorem 2: Value of the levered firm
Under the same assumptions as in theorem 1, the value of the levered firm v(x) is state
dependent and in each state given by:
v1 (x) =
α1 w1 x
for x ≤ K1
α2 λ1 w2 x
B1 xγˆ1 + B2 xγˆ2 + r+λ
+ r+λ
1 −µ1
B3 xβˆ1 + B4 xβˆ2 + w1 x + τ C
(1−τ )x
r+λ1 −µ1
for K1 < x ≤ K2
for K2 < x
v2 (x) =
α2 w2 x
for x ≤ K2
ˆb3 B3 xβˆ1 + ˆb4 B4 xβˆ2 + w2 x +
for K2 < x
where βˆ1 , βˆ2 < 0, γˆ1 > 0, γˆ2 < 0, ˆb3 < 0, ˆb3 > 0, and B1 , B2 , B3 , B4 are the coefficients
derived from the boundary conditions (23) - (26). Explicit formulas of the exponents
and coefficients and the derivation of the formula are contained in the appendix A.2.
The structure of the functions in theorem 2 is identical to those in theorem 1. The value
of the levered firm is the state dependent value of unlevered assets wi x plus the tax shield
τ C. In the action region (K1 < x) the tax shield is independent of the present state, but
the liquidation value incorporates the present state and the switch to the other. In the
transient region the tax shield is discounted at higher rate r + λ1 because a switch from
state one to state two would result in a loss of the tax shield. Again, the negative sum
B3 xβ1 + B4 xβ2 reflects the discount due to default risk and a possible state switch. In state
two the subtraction is larger than in state one due to the higher default risk through the
prevailing negative trend µ2 .
Equity value
The value of a levered firm’s equity is the present value of the residual claim to the levered
firm value after deducing payments to bondholders.
Theorem 3: Equity value
Under the same assumptions as in theorem 1, the value of equity e(x) is state dependent
and given by
ei (x) =
for x ≤ Ki
vi (x) − di (x)
for Ki < x
i = 1, 2 .
In case of liquidation the bondholders receive all that is left of the unlevered firm value
and, hence, equity becomes worthless. Outside of the default region the residual claim is
positive and increasing in EBIT. In the transient region a sudden state switch from state one
to state two results in a total loss for the shareholders. Since the value of corporate debt is
bounded, the growth of the equity value converges to the growth of the firm value for large
x. Debt and levered firm value satisfy smoothness conditions at the upper default boundary
K2 and hence equity does so as well.
Coupon size, default policy and optimal capital structure
In t = 0 the management has to decide about the hight of the principal that will be
issued. The fair coupon C that the firm has to pay for its debt obligations depends on
the principal P , the present state i0 , and the default thresholds K1 , K2 because as show in
theorem 1 and 2 the discount of corporate securities depends on the business cycle. For a
given set of P, x0 , K1 , K2 we can find the fair coupon by solving the debt-at-par equation for
C numerically:
di0 ,C,P,K1 ,K2 (x0 ) = P .
Since management acts in the best interest of shareholders, we employ a smoothness condition according to Leland and Toft (1996) in each state
∂ei (x) =0
∂x x=Ki
i = 1, 2 .
Equation (31) guarantees that default is triggered at the point where marginal increase of
equity value is zero. The value of equity e(x) is an implicit function of the coupon C and
in turn the value of debt d(x) depends on the default thresholds K1 and K2 . When solving
(31) for a given principal P iteratively, C has to be determined in every step by solving (30).
Being able to determine the optimal default thresholds and the fair coupon, we can derive
an optimal capital structure by maximizing the levered firm value. The optimal leverage ratio
is the solution to the problem:
max vi0 (x0 ) s.t. (30), (31)
0<P <vi0
i0 = 1, 2.
Table 1: Summary of the calibrated parameters used in the benchmark case. The parameters are estimated
on aggregated S&P500 date.
growth rate of EBIT
volatility of EBIT
rate of leaving a state λ1
recovery rate
risk-free rate
corporate tax rate
retirement rate of debt m
For each i0 we receive an optimal principal P , a fair coupon C, and two distinguishable
default thresholds K1 and K2 . We denote the solution to (32) with vi∗ (x) and define the
optimal leverages by li∗ = d∗i (x)/vi∗ (x). There are two different optimal capital structures,
one for state one and one for state two.
Implications for optimal capital structure
In this section we focus on the structural estimation of the parameters of the model and
the implication for corporate financing policies. We call the set of estimated parameters
benchmark scenario. Table 1 summarizes the estimated parameters.
Parameter estimation
The parameters λ1 , λ2 that determine the transition of the states of the economy are
estimated to fit stylized facts on the state of the US-economy after the 1960ies. An average
recession lasts for 5 years4 which corresponds to λ1 = 0.2. Being currently in an expansion,
then the probability of entering a recession within one year is about 18% and within two
The average length of state i is given by 1/λi .
years 33%5 . We choose a conservative setting by assuming that the average length of a
recessions is much shorter and set the rate of leaving the recession to λ2 = 0.65. In this
manner a recession lasts on average for 1.54 years and the probablility of leaving a recession
within one year is 48% and 73% within two years. In the long run about 76% of time is
spend in an expansion an 24% in a contraction. Firms’ EBIT follows the positive drift on
average longer than the negative one.
We calibrate the EBIT-process xt to fit the annually aggregated EBIT of S&P500 firms.
We do not use firm level data for calibration because the difference in trends across the
business cycle is more pronounced in aggregated data. An observation year is regarded to
be a recession if at least six month of the fiscal year is considered as a recession by the
monthly NBER recession indicator. Otherwise, the year belongs to an expansion. Using
annual Compustat data from 1962 to 2006, we observe a positive average growth rate of
the aggregated S&P500 EBIT in expansions and a negative growth rate in recessions. Due
to our assumption that investors are risk neutral, we choose µ1 = 0.04 and µ2 = −0.15.
Uncertainty in form of a volatility is intuitively higher in recessions. We choose σ1 = 0.2
and σ2 = 0.25. We set the initial level of EBIT to the arbitrary value x0 = 10. All results in
percent, especially the optimal leverage ratios and the benefit to leverage, do not vary in x0 .
In line with Gilson (1997) and Andrade and Kaplan (1998) who report defaults costs of
20% to 40 %, we choose a recovery rate in an expansion of α1 = 0.80 and in a recession
α2 = 0.60. In line with previous research we set the corporate tax rate τ to 15% and the
risk free rate of interest r to 0.05 approximating the historical average of a short term US
government bond.
Figure 1 shows the path of aggregated S&P 500 EBIT. The graph displays the stylized
facts of the regime switching process. During expansions there is a positive growth in EBIT
while in recession the growth rate is negative. Without the regime switching ability it is
not possible to characterize certain periods as recessions. The possibility of increasing and
The cumulative distribution function of the exponential distrubution Fexp (t) gives the probability that
the event of a state switch occures up to time t: Fexp (t) = 1 − e−λi t .
Figure 1: Aggregated EBIT of S&P500 firms from 1975 to 2001. Recession years classified by NBER are
1981-1982, 1990-1991, and 2001.
decreasing EBIT would be constant throughout time and independent of the business cycle.
Pro-cyclical leverage and benefits
The results of our benchmark scenario in table 2 show that optimal leverage is procyclical. In expansions the firm chooses to finance 50% of their capital needs with debt and
50% to be equity. The positive growth of EBIT last on average 60 month which pushes
EBIT on average further away from the default thresholds and reduces default. When
after an average expansion a state switch to a recession occurs, then, despite the negative
EBIT growth, default is unlikely because the distance between EBIT xt and the default
threshold K2 has increased during the expansion. In contrast, if the present state is already
a recession, then the firm chooses an optimal leverage ratio of 46%. On average EBIT loses
15% continuously within one year, which leads to lower debt capacity and interest coverage.
Management chooses to finance a larger fraction with equity because additonal debt would
increase default risk and reduce the levered firm value.
The benefit to leverage is pro-cyclical. In an expansion the unlevered firm value of 184.02
is increased through debt issuance of 96.23 by 5%. These benefits are obtained in two ways.
First, in the present expansion the firm generates high tax shields and second, when the
next recession enters, the firm has the same amount of debt outstanding but on average at
a higher EBIT level. In this manner the tax benefits remain high in the upcoming recession.
In contrast, if the present state is already a recession, the unlevered asset value of 157.34
is increased through the debt issuance of 72.15 only by 4%. Now, the firm cannot afford
operating at a high leverage ratio and misses tax benefits compared to an expansion. As the
expansion sets in tax benefits become more secure but are much lower compared to those
that could be generated if the present state was already an expansion. Our values roughly
reflect the estimates of benefits to leverage by Graham (2000). Figure 2a shows the levered
firm value in dependence of the chosen leverage. First the levered firm value increase in
both states due to rising tax benefits. At some point the marginal default costs exceed the
additional tax benefits leading to negative net benefits which result in a decreasing firm
value. The point where marginal default costs exceed marginal tax benefits is smaller in a
recession leading to the observed pro-cyclical leverage. The choice of operating at a lower
leverage ratio reflects a more conservative financial policy.
As long as managers choose a principal where the levered firm value exceeds the unlevered
firm value, they create (not necessarily optimal) shareholder value. The points where the
solid and the dashed line hit the dotted lines, are the barrier to the region where managers
destroy shareholder value, because the levered firm is smaller than its unlevered value. This
barrier is pro-cyclical, i.e. in an expansion this point is reached at an leverage of 0.81 and
in an recession at 0.74. Because of the positive growth expectations of EBIT in expansions,
the levered firm value still exceeds the unlevered value at a higher leverage level.
The value of equity is counter-cyclical. This does not imply a higher value for shareholders
in recessions, it rather reflects the choice of financial sources. In our model the costs of
choosing equity financing are lower, but on the other hand there are no benefits in form of
tax shields. The change in shareholders’ wealth is reflected through the unlevered firm value
Table 2: Results on state dependent capital structure using the parameters from the benchmark scenario.
optimal leverage
li ∗
benefit to leverage
levered firm value vi (x0 )
debt value di (x0 )
equity value ei (x0 )
unlevered firm value ui (x0 )
coupon Ci (x0 )
Ratio of default thresholds
and the change in benefits to leverage.
The default thresholds Ki are counter-cyclical as well. Firms default earlier in a recession
than in an expansion due to the negative expected growth rate µ2 . In contrast, if a firm
operates in an expansion at a EBIT level between the two thresholds, it will remain active
because of the expected positive growth of EBIT over time. If a sudden state switch occurs,
then all firms with EBIT in the range of [K1 , K2 ] are liquidated simultaneously, which refers
to the default clustering described by Driessen (2005), Cremers, Driessen, and Maenout
(2008), and Hackbarth, Miao, and Morellec (2006). The counter-cyclical default thresholds
follow from the solution to equation (31) and are observed without initial split up into three
regions in section 2.
Our theoretical implications explain various empirical findings on corporate capital structure. Korteweg (2010) calculates optimal leverage with help of an extended Modigliani and
Miller (1958) setting and observes pro-cyclical leverage ratios and benefits. His empirically
estimated benefits are lower than our predicted values, because our optimal leverage ratios
are based on a risk-neutral setting. Covas and Den Haan (2010) find that debt issuances
are pro-cyclical for most size-sorted US-firms. In contrast, Korajczyk and Levy (2003) state
that leverage ratios are counter-cyclical. These findings do not contradict each other, because Korajczyk and Levy (2003) examine observed leverage ratios which do not correspond
to optimal or target leverage ratios because of market frictions (Leary and Roberts (2005)).
Retained earnings and temporary earnings shock change leverage ratios over time and the
costs of returning to the optimum might exceed the benefits. In our model at the end of
an expansion a firm operates at a leverage ratio that is substantially lower than the optimal
recession-leverage due to the increased EBIT level. Hence, observed leverage ratios appear
to be counter-cyclical while optimal leverage ratios remain pro-cyclical.
In other parsimonious trade-off models of capital structure (Hackbarth, Miao, and Morellec (2006)) optimal leverage is counter-cyclical. Their result is mainly driven by their assumed
EBIT-process that is discontinuous at turning points. Their proposed EBIT process looses
a fraction of its value in recession, but is restored to the old level in the next expansion.
Implicitly the expected growth rate of EBIT in a recession is large because investors expect
a positive jump. As well, in their model the growth rate in expansion is negative, because of
the probability to a switch to a recession. In line with Nieuwerburgh and Veldkamp (2006)
we model a “sharp and short” downturn and a “more gradual” resumption of the expansion
to obtain a realistic setting.
Cost of being over- or underlevered
In this section we show that costs of operating at a non-optimal leverage are higher
in recessions and that capital structure adjustments due to these costs are more common
around turning points of the economy.
Often managers face the problem that they cannot infer the growth rate of their firm’s
EBIT exactly, nor do they know the present state with certainty. If they choose a leverage
ratio based on an estimate that might deviate from the true value, they come to the problem
of operating at a non-optimal leverage. For example, if the manager believes that the present
state is an expansion, but in reality a recession has already started, then his chosen principal
exceeds the optimal value. In this case the firm would be overlevered and face higher default
costs. In contrast, if the firm chooses a principal that is too small, i.e. it underestimates its
growth rate, the firm is underlevered and misses substantial tax benefits. In both cases the
(a) leverage and levered firm value
(b) costs of being over- or underlevered
Figure 2: Panel (a) shows the levered firm value in dependence of chosen leverage. All parameters correspond to the values in table 1. The solid line is the levered firm value in a expansion (i = 1), the dashed line
in a recession (i = 2). The upper dotted line is the unlevered firm value in an expansion and the lower dotted
line the unlevered firm value in a recession. The optimal levered firm values are marked with a circle. Panel
(b) shows the costs of being over- or underlevered. The x-axis is the deviation D of the chosen principal from
the optimal leverage in the corresponding state. The cost of being over- or underlevered ci (D) are measured
as difference between the optimal and the chosen levered firm value in percent. The solid lines are the cost
in an expansion and the dashed line in a recession.
levered firm value is smaller than the optimum.
We measure the costs of being over- or underlevered ci as the loss in the levered firm
value as percentage of optimal value vi∗ and dependent on the present state i = 1, 2. In this
context we regard the levered firm value vi as a function of the (non-optimal) principal P and
define the difference between chosen principal P and optimal principal P ∗ as D = P − P ∗ .
Initial EBIT x0 are treated as a constant parameter of the levered firm value. We can now
write the costs of being over- or underlevered ci (D) as
ci (D) = 1 −
vi (P ∗ + D)
vi (P ∗ )
Figure 2b plots the costs ci against the difference in principals D for both states i. For D = 0
the firm incurs no additional costs. If the principal is too low, i.e. D < 0, the firm faces
costs of being underlevered due to missing tax benefits because the tax deductible coupon
payments are lower than they are in the optimal case. These costs are higher in recessions
since expected future earnings decrease on average and other means of creating shareholder
value become more important. The costs of being underlevered are limited to maximum
benefits to leverage.
For D > 0 the firm faces costs of being overlevered. Beyond the optimal principal
marginal default costs increase more rapidly than tax benefits due to the high default probability. The difference in growth rates across the business cycle leads to higher costs of being
overlevered in recessions. The loss of shareholder value is in this case not limited to the tax
benefits because a very high leverage can diminish the levered value to the liquidation value
of the assets. Combining our results, our model shows that in recessions it is more important
to operate at the optimal leverage because additional costs can reduce the shareholder value
more heavily.
For a small deviation from the optimum the costs do not exceed a high percentage of
the levered firm value. For example, adjustment costs for debt-equity swaps to return to
the optimum might outweigh the benefits. If adjustment costs are smaller than 1% of the
levered firm value, then a deviation of -40 or +30 units of debt would be still be less costly
than adjusting to the optimum. Hence, firms do not tend to adjust their leverage often as
long as a state switch does not occur and the optimum remains the same. In contrast, if
a state switch occures then the optimal principal moves by about 24 units. Especially a
switch from an expanison to a recession leads to costs of overlevering, because the optimal
principal is reduced by those 24 units. Now, the costs of being overlevered can exceed some
percentage points of the levered firm value and it would be optimal to adjust to the target
leverage. Therefore, leverage adjustments should be more common around turning points
than within a state of the business cycle. Since costs of being overlevered are higher in
recessions, adjustments are more likely in this state of the economy than in expansions.
Our theoretical results establish empirically testable implications on the timing of capital
structure adjustments.
Table 3: Results on the parameter estimation for changing macroeconomic conditions. All parameters but
µ2 , σ2 , and λ2 are held constant to the values of the benchmark scenario. The bold number indicate the
benchmark scenario.
v1 (x0 ) v2 (x0 )
174.9 138.8
192.9 157.3
216.2 181.6
261.0 228.1
192.7 157.1
192.9 157.3
193.2 157.8
193.6 158.2
145.3 108.0
192.9 157.3
235.5 201.6
273.7 241.4
Changes in macroeconomic risk
As noted by Stock and Watson (2002) macroeconomic risk has changed over time. A
shift in macroeconomic conditions influences the market value of corporate debt and equity
because it strongly affects the default risk. In our model there are three ways to reflect
macroeconomic risk: the size of the negative growth rate µ2 , the volatility of EBIT in
recessions σ2 and the rate of leaving a recession λ2 .
First, we analyze the impact of the growth rate in contractions on optimal leverage and
the benefit to leverage. Table 3 shows the results of a variation in µ2 from -0.18 to -0.03. As
macroeconomic risk decreases, the optimal leverage ratios increase by 7% and benefits rise
by 1.5% in both states. The effect of a changing growth rate is reflected the strongest in
the levered firm value which rises by 50% and is driven by an increase of the unlevered asset
value that rises due to the change in µ2 . The firm responds to lower risk with an increase in
leverage that comes with the trade-off of a higher default probability. In comparison to the
levered firm value the change in benefits to leverage is only marginal.
As a second approach, we focus on the effect of a change in the EBIT volatility. σ2 reflects
the deviation from the downtrend of EBIT and is twofold. The chances of achieving high
EBIT levels increase at the cost of a higher default probability. The results of our parameter
variation in table 3 yield that the downside risk is larger because the firm operates at a lower
leverage ratio for high values of σ2 . The decline in the optimal leverage ratio in recessions is
more sharply than in expansions because a present recession influences the valuation stronger
than future recessions due to discounting of cash flows. The benefit to leverage decreases
slightly from 5.2% (5%) to 4.7% (4.2%) as volatility increases because the firm is less likely
to benefit from future tax shields. In contrast, the levered firm value is almost unaffected by
an increase in σ2 which indicates that the change in the tax shield and in the default costs
are of equal magnitude. As shown in equation (8), the unlevered firm value is independent
of σ2 and constant in the parameter variation.
The third factor that characterizes a recession is the expected length of the state which
equals the inverse of the rate of leaving state two λ2 . The length of a contraction determines
the time during which the EBIT is exposed to the negative growth rate µ2 . Our results
yield that a short recession (i.e. a high rate of leaving a recession) results in higher optimal
leverage ratios (0.57 and 0.54). In case of a long recession the optimal ratios are much lower
(0.44 and 0.39). The levered firm value is strongly affected by the length of a recession and
so are the benefits to leverage that almost double.
Combining the results of the three parameter variations, we find that a reduction in
macroeconomic risk increases optimal leverage. By acting optimally firms can achieve similar
benefits by adjusting their financial policy to the new conditions. Shareholders profit primary
through the increase in the unlevered firm value, but not through the benefits to debt
financing which remain almost equal. Applying this finding to corporate financing policies,
we conclude that after the observed change of macroeconomic risk in the 1980s, debt became
more attractive to firms because the default risk was reduced. Speaking empirically, optimal
leverage ratio should be higher after the change in the macroeconomic risk.
Concluding remarks
Our study shows that macroeconomic conditions are an important determinant of capital
structure decisions. Optimal leverage and benefits to leverage depend on the present state
of the economy and on the additional risk of a sudden switch to the other state. Optimal
leverage ratios and benefits to leverage vary pro-cyclically, i.e. they are higher in expansions,
because of varying growth rates of EBIT. Even though debt becomes more risky in recessions,
tax benefits remain an important mean to maximize the levered firm value.
A change in macroeconomic risk, such as milder or less volatile recessions, leads to an
increase of the levered firm value and the optimal leverage ratios. However, benefits to
leverage increase only a little. Hence, after the change in macroeconomic risk in the 1980s,
firm values and leverage ratios should have gone up, but firms should not have profited much
from increasing tax benefits.
Our model shows that being over- or underlevered is more costly in recessions and that
capital structure adjustments due to these costs are more common in recessions or around
turning points. If a firm has problems to determine its optimal capital structure exactly,
then the firm should act more conservatively and issue less debt. The costs of the lower tax
shield are smaller than possible high default costs.
Our static setting can be extended to a dynamic setting, that would give the firm the
option of restructuring its capital if EBIT have reached an upper threshold. However, the
results of Hackbarth, Miao, and Morellec (2006) and Bhamra, Kuehn, and Strebulaev (2009)
indicate that dynamics in structural models do not change the order of the results, only
the level of leverage. Hackbarth, Miao, and Morellec (2006) find that optimal leverage is
smaller when the firm has the option to issue debt in the future, but the cyclicality of debt
issuances remains. In our model we omit the option of future debt issuance to keep the
model parsimonious.
Andrade, G., and S. Kaplan (1998): “How Costly Is Financial (Not Economic) Distress? Evidence from Highly Leveraged Transactions That Became Distressed,” Journal
of Finance, 53, 1443–1493.
Bhamra, H. S., L.-A. Kuehn, and I. A. Strebulaev (2009): “The Aggregate Dynamics
of Capital Structure and Macroeconomic Risk,” Review of Financial Studies, 23(2).
Chen, H. (2010): “Macroeconomic Conditions and the Puzzles of Credit Spreads and Capital Structure,” Journal of Finance, 65, 2171–2212.
Covas, F., and W. J. Den Haan (2010): “The Cyclical Behavior of Debt and Equity
Finance,” American Economic Review, Forthcoming.
Cremers, M., J. Driessen, and P. Maenout (2008): “Explaining the Level of Credit
Spreads: Option-Implied Jump Risk Premia in a Firm Value Model,” Review of Financial
Studies, 21, 2209–2242.
Driessen, J. (2005): “Is Default event risk priced in corporate bonds?,” Review of Financial
Studies, 18, 165–195.
Driffill, J., M. Raybaudi, and M. Sola (2003): “Investment under uncertainty with
stochastically switching Profit Streams,” Studies in Nonlinear Dynamics & Econometrics,
7(1), 1–38.
Gilson, S. (1997): “Transaction Costs and Capital Structure Choice: Evidence from Financially Distressed Firms,” Journal of Finance, 52, 161–196.
Goldstein, R., N. Ju, and H. E. Leland (2001): “An EBIT-Based Model of Dynamic
Capital Structure,” The Journal of Business, 74(4), 483–512.
Graham, J. R. (2000): “How Big Are the Tax Benefits of Debt,” Journal of Finance, 55,
Guo, X. (2001): “Information and option pricings,” Quantitive Finance, 1, 38–44.
Hackbarth, D., J. Miao, and E. Morellec (2006): “Capital Structure, Credit Risk,
and Macroeconomic Conditions,” Journal of Financial Economics, 82(3), 519–550.
Korajczyk, R., and A. Levy (2003): “Capital structure choice: macroeconomic conditions and financial constraints,” Journal of Financial Economics, 68, 75–109.
Korteweg, A. (2010): “The Net Benefits to Leverage,” Journal of Finance, 65, 2137–2170.
Leary, M., and M. Roberts (2005): “Do firms rebalance their capital structures?,”
Journal of Finance, 60, 2575–2619.
Leland, H. E. (1994): “Corporate Debt Value Bond Covenants and Optimal Capital
Structure,” Journal of Finance, 49(4), 1213–1252.
(1998): “Agency Costs, Risk Management, and Capital Structure,” The Journal
of Finance, 53, 1213–1243.
Leland, H. E., and K. B. Toft (1996): “Optimal Capital Structure, Endogenous
Bankruptcy, and the Term Structure of Credit Spreads,” The Journal of Finance, 51(3),
Mello, A., and J. Parsons (1992): “Measuring the agency cost of debt,” Journal of
Finance, 47(5), 1887–1904.
Modigliani, F., and M. Miller (1958): “The Cost of Capital, Corporation Finance, and
the Theory of Investment,” American Economic Review, 48, 261–297.
Nieuwerburgh, S. V., and L. Veldkamp (2006): “Learning asymmetries in real business
cycles,” Journal of Monetary Economics, 53(4), 753–772.
Stock, J., and M. Watson (2002): “Has the Business Cycled Changed and Why?,”
NBER Macroeconomic Annual, 53, 159–218.
Derivation of the differential equations in section
The derivation of the differential equations for the levered firm, the unlevered firm, and
debt value follows Driffill, Raybaudi, and Sola (2003). First, we regard the unlevered firm
value ut = u(xt ) of the firm which corresponds to the value of a firm that does not issue
any debt. Let E denote the expectation operator on the σ-algebra that is generated by the
Brownian motion and the Markov chain, and EW denotes the expectation operator on the
σ-algebra of just the Brownian motion. The infinitesimal change of its value can be described
by the following equation:
rut dt = (1 − τ )xt dt + E[dut ] .
Since all investors are risk neutral, all cash flows are discounted at the risk free rate r. The
required rate of return r equals the growth of after tax EBIT (1 − τ )xt plus the change in its
value E[du] that arises because of a variation of xt and a possible state switch. Let ui denote
the unlevered asset value conditioned on the state i. We assume that the present state is 1,
so i = 1. With the transition probability λ1 we have
E[ut+dt ] = (1 − λ1 dt) EW [u1 (xt+dt )] + λ1 dt EW [u2 (xt+dt )].
The first summand denotes the event of remaining in state 1 times its probability, and the
second part is the case of switching to state 2. The expected change of U on the interval dt
E[du] = E[u(xt+dt ) − u(xt )]
= (1 − λ1 dt) EW [u1 (xt+dt )] + λ1 dt EW [u2 (xt+dt )] − EW [u1 (xt )]
= EW [u1 (xt+dt ) − u1 (xt )] + λ1 dt EW [u2 (xt+dt ) − u1 (xt+dt )]
The second summand describes a switch from state one to state two and is independent of
the expectation operator EW . The first part equals the change of u given that the economy
remains in the first state. Under the assumption of remaining in state one xt is an Itˆo-process,
so that we can apply Itˆo’s lemma to u1 (xt ) and receive
1 2 2 00
du1 = µ1 xu1 + σ x u1 dt + σ1 xu01 dWt .
It holds that EW (σ1 x dWt ) = 0 and we receive
1 2 2 00
E[du] = µ1 xu1 + σ1 x u1 dt + λ1 (u2 − u1 ) dt .
Using (38), equation (34) equals
ru dt = (1 − τ )x dt +
µ1 xu01
+ σ12 x2 u001
dt + λ1 (u2 − u1 ) dt .
Hence, the differential equation to determine u1 is
1 2 2 00
x σ1 u1 + µ1 xu01 − ru1 + λ1 (u2 − u1 ) + (1 − τ )x = 0 .
If we assume that the present state is 2, the same application of Itˆo’s lemma yields
1 2 2 00
x σ2 u2 + µ2 xu02 − ru2 + λ1 (u1 − u2 ) + (1 − τ )x = 0 .
(5) and (6) form a system of ordinary differential equations that describe the liquidation
value of the firm.
In order to derive differential equations for the value of corporate debt and the levered
firm, one needs to substitute the continuous payment (1 − τ )x in (34) by C + mP or (1 −
τ )x + τ C, respectively.
In the case of corporate debt value di (x) the firm defaults in state 2 but remains active
in state 1 if K1 < xt < K2 . (35) reads in this case:
E[dt+dt ] = (1 − λ1 dt) EW [d1 (xt+dt )] + λ1 dt EW [α2 u2 (xt+dt )],
where α2 is the recovery rate in state 2. With this modification one can derive differential
equations for the transient region by applying the same exercise as above.
Solving the differential equations
We will consider the case of corporate debt only. In the transient region we have to solve
equation (11):
1 2 2 00
x σ1 d1 + µ1 xd01 − (r + m)d1 + λ1 (α2 w2 x − d1 ) + C + mP = 0
which is a linear ordinary differential equation of second order. The homogeneous part of
the solution reads
(x) = A1 xγ + A2 xγ ,
where γi are the roots of the characteristic equation of the equivalent differential equation
with constant coefficients. γi satisfies
= 0.5 − 2 ±
0.5 − 2
r + m + λ1
The particular solution of (11) reads
1 (x) =
C + mP
α 2 λ1 w 2 x
r + m + λ1 r + m + λ1 − µ1
w2 is the constant coefficient for the function for the value of unlevered assets ui (x) = wi x.
Combined, we have a function with two unknown coefficients A1 and A2 :
d1 (x) = A1 xγ1 + A2 xγ2 +
α2 λ1 w2 x
C + mP
r + m + λ1 r + m + λ1 − µ1
In the action region the system of differential equations (12) and (13)
1 2 2 00
x σ1 d1 + µ1 xd01 − (r + m)d1 + λ1 (d2 − d1 ) + C + mP = 0
1 2 2 00
x σ2 d2 + µ2 xd02 − (r + m)d2 + λ2 (d1 − d2 ) + C + mP = 0
is linear and of second order as well. By transforming the equations to a system with constant
coefficients, one obtains the homogeneous solution functions
(x) = A3 xβ1 + A4 xβ2 + A7 xβ3 + A8 xβ4 ,
(x) = A5 xβ1 + A6 xβ2 + A9 xβ3 + A10 xβ4 .
βi are the four distinct roots of the fourth order polynomial
−a1 1 − a4 − β −a3
det 
 0
−a2 1 − a5 − β
a1 =
a4 =
a2 =
a5 =
a3 =
a6 =
that have to be calculated numerically. The particular solutions to (12) and (13) read
(x) =
C + mP
i = 1, 2.
Combining (45), (45), and (49), the general solution functions have eight unknown coefficients:
C + mP
C + mP
d1 (x) = A3 xβ1 + A4 xβ2 + A7 xβ3 + A8 xβ4 +
d2 (x) = A5 xβ1 + A6 xβ2 + A9 xβ3 + A10 xβ4
In line with Driffill, Raybaudi, and Sola (2003) we find that β1, β2 < 0 and β3, β4 > 0.
Since the solution functions are bounded by the value of risk-free debt
it follows that
A7 = A8 = A9 = A10 = 0. Combining (44), (50), and (51) there are six coefficients left
A1 , A2 , A3 , A4 , A5 , A6 to be determined.
For every pair of real numbers Aj , Ai there is another number bj that satisfies bj Aj = Ai .
In our case we can determine bj independent of x and hence constant for a given set of
parameters. Assuming b5 A5 = A3 and b6 A6 = b4 and plugging (50) and (51) into the
differential equation (12) leads to
1 1 2
b5 = −
σ β1 (β1 − 1) + µ1 β1 − (r + m) − λ1
λ1 2 1
1 1 2
b6 = −
σ β2 (β2 − 1) + µ1 β2 − (r + m) − λ1
λ1 2 1
Now, there are four unknown coefficients left that can be determined uniquely with the
boundary conditions (14) - (17).
When deriving the solutions for the value of the levered firm vi (x), one needs to set
m = 0 and use the boundary conditions (23) - (26). The coefficients of the terms with
positive exponents equal to zero because the levered firm value is bounded by wi x + τ C. βˆi
and γˆi correspond to βi and γi with m = 0, i = 1, 2.
Centre for Financial Research
CFR Working Paper Series
CFR Working Papers are available for download from
Hardcopies can be ordered from: Centre for Financial Research (CFR),
Albertus Magnus Platz, 50923 Koeln, Germany.
D. Hess, P. Immenkötter
Optimal Leverage, its Benefits, and the Business Cycle
N. Heinrichs, D. Hess,
C. Homburg, M. Lorenz,
S. Sievers
Extended Dividend, Cash Flow and Residual Income Valuation
Models – Accounting for Deviations from Ideal Conditions
A. Kempf, O. Korn,
S. Saßning
Portfolio Optimization using Forward - Looking Information
V. Agarwal, S. Ray
Determinants and Implications of Fee Changes in the Hedge
Fund Industry
G. Cici, L.-F. Palacios
On the Use of Options by Mutual Funds: Do They Know What
They Are Doing?
V. Agarwal, G. D. Gay,
L. Ling
Window Dressing in Mutual Funds
N. Hautsch, D. Hess,
D. Veredas
The Impact of Macroeconomic News on Quote Adjustments,
Noise, and Informational Volatility
G. Cici
The Prevalence of the Disposition Effect in Mutual Funds'
S. Jank
Mutual Fund Flows, Expected Returns and the Real Economy
G.Fellner, E.Theissen
Short Sale Constraints, Divergence of Opinion and Asset
Value: Evidence from the Laboratory
Are There Disadvantaged Clienteles in Mutual Funds?
V. Agarwal, C. Meneghetti The Role of Hedge Funds as Primary Lenders
G. Cici, S. Gibson,
J.J. Merrick Jr.
Missing the Marks? Dispersion in Corporate Bond Valuations
Across Mutual Funds
J. Hengelbrock,
Market Response to Investor Sentiment
E. Theissen, C. Westheide
G. Cici, S. Gibson
The Performance of Corporate-Bond Mutual Funds:
Evidence Based on Security-Level Holdings
D. Hess, D. Kreutzmann,
O. Pucker
Projected Earnings Accuracy and the Profitability of Stock
S. Jank, M. Wedow
Sturm und Drang in Money Market Funds: When Money
Market Funds Cease to Be Narrow
G. Cici, A. Kempf, A.
Caught in the Act:
How Hedge Funds Manipulate their Equity Positions
J. Grammig, S. Jank
Creative Destruction and Asset Prices
S. Jank, M. Wedow
Purchase and Redemption Decisions of Mutual Fund
Investors and the Role of Fund Families
S. Artmann, P. Finter,
A. Kempf, S. Koch,
E. Theissen
The Cross-Section of German Stock Returns:
New Data and New Evidence
M. Chesney, A. Kempf
The Value of Tradeability
S. Frey, P. Herbst
The Influence of Buy-side Analysts on
Mutual Fund Trading
V. Agarwal, W. Jiang,
Y. Tang, B. Yang
Uncovering Hedge Fund Skill from the Portfolio Holdings They
V. Agarwal, V. Fos,
W. Jiang
Inferring Reporting Biases in Hedge Fund Databases from
Hedge Fund Equity Holdings
V. Agarwal, G. Bakshi,
J. Huij
Do Higher-Moment Equity Risks Explain Hedge Fund
J. Grammig, F. J. Peter
Tell-Tale Tails
K. Drachter, A. Kempf
Höhe, Struktur und Determinanten der ManagervergütungEine Analyse der Fondsbranche in Deutschland
J. Fang, A. Kempf,
M. Trapp
Fund Manager Allocation
P. Finter, A. NiessenRuenzi, S. Ruenzi
The Impact of Investor Sentiment on the German Stock Market
D. Hunter, E. Kandel,
S. Kandel, R. Wermers
Endogenous Benchmarks
S. Artmann, P. Finter,
A. Kempf
Determinants of Expected Stock Returns: Large Sample
Evidence from the German Market
E. Theissen
Price Discovery in Spot and Futures Markets:
A Reconsideration
M. Trapp
Trading the Bond-CDS Basis – The Role of Credit Risk
and Liquidity
A. Kempf, O. Korn,
M. Uhrig-Homburg
The Term Structure of Illiquidity Premia
W. Bühler, M. Trapp
Time-Varying Credit Risk and Liquidity Premia in Bond and
CDS Markets
W. Bühler, M. Trapp
Explaining the Bond-CDS Basis – The Role of Credit Risk and
S. J. Taylor, P. K. Yadav,
Y. Zhang
Cross-sectional analysis of risk-neutral skewness
A. Kempf, C. Merkle,
A. Niessen
Low Risk and High Return - How Emotions Shape
Expectations on the Stock Market
V. Fotak, V. Raman,
P. K. Yadav
Naked Short Selling: The Emperor`s New Clothes?
F. Bardong, S.M. Bartram, Informed Trading, Information Asymmetry and Pricing of
P.K. Yadav
Information Risk: Empirical Evidence from the NYSE
S. J. Taylor , P. K. Yadav,
Y. Zhang
The information content of implied volatilities and model-free
volatility expectations: Evidence from options written on
individual stocks
S. Frey, P. Sandas
The Impact of Iceberg Orders in Limit Order Books
H. Beltran-Lopez, P. Giot,
J. Grammig
Commonalities in the Order Book
J. Fang, S. Ruenzi
Rapid Trading bei deutschen Aktienfonds:
Evidenz aus einer großen deutschen Fondsgesellschaft
A. Banegas, B. Gillen,
A. Timmermann,
R. Wermers
The Performance of European Equity Mutual Funds
J. Grammig, A. Schrimpf,
M. Schuppli
Long-Horizon Consumption Risk and the Cross-Section
of Returns: New Tests and International Evidence
O. Korn, P. Koziol
The Term Structure of Currency Hedge Ratios
U. Bonenkamp,
C. Homburg, A. Kempf
Fundamental Information in Technical Trading Strategies
O. Korn
Risk Management with Default-risky Forwards
J. Grammig, F.J. Peter
International Price Discovery in the Presence
of Market Microstructure Effects
C. M. Kuhnen, A. Niessen
Public Opinion and Executive Compensation
A. Pütz, S. Ruenzi
Overconfidence among Professional Investors: Evidence from
Mutual Fund Managers
P. Osthoff
What matters to SRI investors?
A. Betzer, E. Theissen
Sooner Or Later: Delays in Trade Reporting by Corporate
P. Linge, E. Theissen
Determinanten der Aktionärspräsenz auf
Hauptversammlungen deutscher Aktiengesellschaften
N. Hautsch, D. Hess,
C. Müller
Price Adjustment to News with Uncertain Precision
D. Hess, H. Huang,
A. Niessen
How Do Commodity Futures Respond to Macroeconomic
R. Chakrabarti,
W. Megginson, P. Yadav
Corporate Governance in India
C. Andres, E. Theissen
Setting a Fox to Keep the Geese - Does the Comply-or-Explain
Principle Work?
M. Bär, A. Niessen,
S. Ruenzi
The Impact of Work Group Diversity on Performance:
Large Sample Evidence from the Mutual Fund Industry
A. Niessen, S. Ruenzi
Political Connectedness and Firm Performance:
Evidence From Germany
O. Korn
Hedging Price Risk when Payment Dates are Uncertain
A. Kempf, P. Osthoff
SRI Funds: Nomen est Omen
J. Grammig, E. Theissen,
O. Wuensche
Time and Price Impact of a Trade: A Structural Approach
V. Agarwal, J. R. Kale
On the Relative Performance of Multi-Strategy and Funds of
Hedge Funds
M. Kasch-Haroutounian,
E. Theissen
Competition Between Exchanges: Euronext versus Xetra
V. Agarwal, N. D. Daniel,
N. Y. Naik
Do hedge funds manage their reported returns?
N. C. Brown, K. D. Wei,
R. Wermers
Analyst Recommendations, Mutual Fund Herding, and
Overreaction in Stock Prices
A. Betzer, E. Theissen
Insider Trading and Corporate Governance:
The Case of Germany
V. Agarwal, L. Wang
Transaction Costs and Value Premium
J. Grammig, A. Schrimpf
Asset Pricing with a Reference Level of Consumption:
New Evidence from the Cross-Section of Stock Returns
V. Agarwal, N.M. Boyson,
N.Y. Naik
Hedge Funds for retail investors?
An examination of hedged mutual funds
D. Hess, A. Niessen
The Early News Catches the Attention:
On the Relative Price Impact of Similar Economic Indicators
A. Kempf, S. Ruenzi,
T. Thiele
Employment Risk, Compensation Incentives and Managerial
Risk Taking - Evidence from the Mutual Fund Industry -
M. Hagemeister, A. Kempf CAPM und erwartete Renditen: Eine Untersuchung auf Basis
der Erwartung von Marktteilnehmern
S. Čeljo-Hörhager,
A. Niessen
How do Self-fulfilling Prophecies affect Financial Ratings? - An
experimental study
R. Wermers, Y. Wu,
J. Zechner
Portfolio Performance, Discount Dynamics, and the Turnover
of Closed-End Fund Managers
U. v. Lilienfeld-Toal,
S. Ruenzi
A. Kempf, P. Osthoff
Why Managers Hold Shares of Their Firm: An Empirical
The Effect of Socially Responsible Investing on Portfolio
R. Wermers, T. Yao,
J. Zhao
The Investment Value of Mutual Fund Portfolio Disclosure
M. Hoffmann, B. Kempa
The Poole Analysis in the New Open Economy
Macroeconomic Framework
K. Drachter, A. Kempf,
M. Wagner
Decision Processes in German Mutual Fund Companies:
Evidence from a Telephone Survey
J.P. Krahnen, F.A.
Schmid, E. Theissen
Investment Performance and Market Share: A Study of the
German Mutual Fund Industry
S. Ber, S. Ruenzi
On the Usability of Synthetic Measures of Mutual Fund NetFlows
A. Kempf, D. Mayston
Liquidity Commonality Beyond Best Prices
O. Korn, C. Koziol
Bond Portfolio Optimization: A Risk-Return Approach
O. Scaillet, L. Barras, R.
False Discoveries in Mutual Fund Performance: Measuring
Luck in Estimated Alphas
A. Niessen, S. Ruenzi
Sex Matters: Gender Differences in a Professional Setting
E. Theissen
An Analysis of Private Investors´ Stock Market Return
T. Foucault, S. Moinas,
E. Theissen
Does Anonymity Matter in Electronic Limit Order Markets
R. Kosowski,
A. Timmermann,
R. Wermers, H. White
Can Mutual Fund „Stars“ Really Pick Stocks?
New Evidence from a Bootstrap Analysis
D. Avramov, R. Wermers
Investing in Mutual Funds when Returns are Predictable
K. Griese, A. Kempf
Liquiditätsdynamik am deutschen Aktienmarkt
S. Ber, A. Kempf,
S. Ruenzi
Determinanten der Mittelzuflüsse bei deutschen Aktienfonds
M. Bär, A. Kempf,
S. Ruenzi
Is a Team Different From the Sum of Its Parts?
Evidence from Mutual Fund Managers
M. Hoffmann
Saving, Investment and the Net Foreign Asset Position
S. Ruenzi
Mutual Fund Growth in Standard and Specialist Market
A. Kempf, S. Ruenzi
Status Quo Bias and the Number of Alternatives - An Empirical
Illustration from the Mutual Fund Industry
J. Grammig, E. Theissen
Is Best Really Better? Internalization of Orders in an Open
Limit Order Book
H. Beltran, J. Grammig,
A.J. Menkveld
Understanding the Limit Order Book: Conditioning on Trade
M. Hoffmann
Compensating Wages under different Exchange rate Regimes
M. Hoffmann
Fixed versus Flexible Exchange Rates: Evidence from
Developing Countries
A. Kempf, C. Memmel
On the Estimation of the Global Minimum Variance Portfolio
S. Frey, J. Grammig
Liquidity supply and adverse selection in a pure limit order
book market
N. Hautsch, D. Hess
Bayesian Learning in Financial Markets – Testing for the
Relevance of Information Precision in Price Discovery
A. Kempf, K. Kreuzberg
Portfolio Disclosure, Portfolio Selection and Mutual Fund
Performance Evaluation
N.F. Carline, S.C. Linn,
P.K. Yadav
Operating performance changes associated with corporate
mergers and the role of corporate governance
J.J. Merrick, Jr., N.Y. Naik, Strategic Trading Behaviour and Price Distortion in a
P.K. Yadav
Manipulated Market: Anatomy of a Squeeze
N.Y. Naik, P.K. Yadav
Trading Costs of Public Investors with Obligatory and
Voluntary Market-Making: Evidence from Market Reforms
A. Kempf, S. Ruenzi
Family Matters: Rankings Within Fund Families and
Fund Inflows
V. Agarwal, N.D. Daniel,
N.Y. Naik
Role of Managerial Incentives and Discretion in Hedge Fund
V. Agarwal, W.H. Fung,
J.C. Loon, N.Y. Naik
Risk and Return in Convertible Arbitrage:
Evidence from the Convertible Bond Market
A. Kempf, S. Ruenzi
Tournaments in Mutual Fund Families
I. Chowdhury, M.
Hoffmann, A. Schabert
Inflation Dynamics and the Cost Channel of Monetary
Cfr/University of cologne
D-50923 Cologne
Fon +49(0)221-470-6995
Fax +49(0)221-470-3992
[email protected]